Is raising the school leaving age enough to decrease dropping out?

This paper examines the effects of increasing the compulsory school leaving age from 16 to 18 in Hungary using a difference-in-regression-discontinuities design identification strategy. While the reform increased the length of schooling, it did not decrease the probability of dropping out of secondary school, either on average or among the most at-risk group of Roma ethnic minority young people. Due to grade retentions, marginal students were older than their peers and couldn't have made it to the final grade of secondary school by age 18. Neither did the reform increase the probability of employment at age 20 and 25. I show descriptive evidence that the share of disadvantaged students increased heavily in vocational training schools -that most marginal students attended- and potentially crowded out resources. This mechanism raises concerns about using school leaving age reforms as instrumental variables for education as it may violate its monotonicity assumption.